Helminth Infection Protects Mice from Anaphylaxis via IL-10-Producing B Cells1

Modulation of the immune system by infection with helminth parasites, including schistosomes, is proposed to reduce the levels of allergic responses in infected individuals. In this study we investigated whether experimental infection with Schistosoma mansoni could alter the susceptibility of mice to an extreme allergic response, anaphylaxis. We formally demonstrate that S. mansoni infection protects mice from an experimental model of systemic fatal anaphylaxis. The worm stage of infection is shown to mediate this protective effect. In vivo depletion studies demonstrated an imperative role for B cells and IL-10 in worm-mediated protection. Furthermore, worm infection of mice increases the frequency of IL-10-producing B cells compared with that in uninfected mice. However, transfer of B cells from worm-infected mice or in vitro worm-modulated B cells to sensitized recipients exacerbated anaphylaxis, which was attributed to the presence of elevated levels of IL-4-producing B cells. Worm-modulated, IL-10-producing B cells from IL-4-deficient, but not IL-5-, IL-9- or IL-13-deficient, mice conferred complete resistance to anaphylaxis when transferred to naive mice. Therefore, we have dissected a novel immunomodulatory mechanism induced by S. mansoni worms that is dependent on an IL-10-producing B cell population that can protect against allergic hypersensitivity. These data support a role for helminth immune modulation in the hygiene hypothesis and further illustrate the delicate balance between parasite induction of protective regulatory (IL-10) responses and detrimental (IL-4) allergic responses.

T he frequency and severity of allergic disorders are steadily increasing, particularly in developing countries where ϳ20 -30% of the population are atopic (1). Atopic individuals may develop a spectrum of allergic diseases, including asthma, allergic rhinoconjunctivitis, atopic dermatitis, food allergies, and anaphylaxis. Mechanistically, atopic conditions induce a type 2 cytokine-polarizing response that is characterized by increased Th2 cell development and production of IL-4, IL-5, IL-9, and IL-13, resulting in IgE production and eosinophilia (2).
The reasons for increases in allergic diseases are multifactorial, but concurrent improvement in sanitation and reduction in childhood infections in developed countries have led to the speculation that early infections may reduce the risk of developing allergy (3,4). Paradoxically, in the context of this model, the so-called hygiene hypothesis, human populations with high rates of parasitic helminth infections, which induce an immunological shift toward Th2 responses, have a lower prevalence of allergic disorders (5). Schistosoma spp. are tropical helminth parasites that have been proposed to ameliorate atopic disorders in humans (5). Schistosomes are characteristically associated with being potent inducers of Th2 cytokine responses, including eosinophilia and IgE responses. Despite type 2 responses being considered necessary for the development of allergies, the presence of schistosome infections in humans may reduce allergic responses in infected popu-lations (6). Thus, Schistosoma hematobium-infected school children in Gabon had a lower prevalence of skin reactivity to house dust mites than those free of this infection (7). Therefore, the chronic down-regulation of the immune system during helminth infection evokes a regulatory environment that may impart protection from allergies (8).
It is well characterized that the propensity of schistosomes to induce Th2 cytokines is largely egg-mediated (9). We have previously shown that mice that overexpress IL-13 develop anaphylaxis after sensitization and re-exposure to Schistosoma mansoni eggs (10). Anaphylaxis is an acute, life-threatening, allergic reaction that can cause fatalities after exposure to allergenic Ags in sensitized atopic individuals. Being an immediate-type hypersensitivity reaction, anaphylaxis is characterized by an overproduction of Th2 cytokines contributing to elevated levels of IgE and eosinophilia. Classically, anaphylaxis was considered to involve allergen cross-linking of IgE bound to its high affinity receptor, Fc⑀R1, on mast cells, initiating an intracellular signal transduction cascade that culminates in cell degranulation and the release of potent systemic mediators, including histamine and platelet-activating factor (PAF), 3 which can induce acute bronchoconstriction, tachycardia, hypotension, vascular leakage, and ultimately death (11). In earlier studies, schistosome egg sensitization induced all of these characteristic features in IL-13 transgenic mice, with injection of S. mansoni eggs inducing a rapid and highly elevated production of Th2 cytokines and Ag-specific IgE, with subsequent exposure to eggs inducing fatal anaphylaxis with attendant mast cell degranulation and histamine release (10). However, IgE does not account for all anaphylactic pathways because mice deficient in IgE can develop systemic anaphylaxis (12). Indeed, a recent study has identified a distinct role in mice for alternative mechanisms of anaphylaxis involving IgG, Fc␥RIII, and macrophages (11).
Unexpectedly, despite IL-13 transgenic mice being predisposed to anaphylaxis caused by schistosome eggs (10), these mice had no anaphylactic symptoms during acute or chronic S. mansoni infection. Therefore, in this study we evaluated whether schistosome infection renders mice less susceptible to anaphylaxis using an experimental murine model of systemic anaphylaxis. We have identified that the worm stage of S. mansoni infection modulates mice so they are completely refractory to active or passive systemic anaphylaxis. We now demonstrate that S. mansoni worms up-regulate IL-10-producing B cells that can protect against anaphylaxis.

Parasitology
A Puerto Rican strain of S. mansoni was maintained by passage in Tyler's original strain mice and albino Biomphalaria glabrata snails. Mice were infected with 350 cercariae percutaneously to establish worm and egg infections for Ag production. Six-to 8-wk-old female BALB/c mice were infected percutaneously with 40 male cercariae for experimental unisexual (worm-only) infections and with 40 male and female cercariae for bisexual (wormϩegg) infections. Unisexual male-only snail infections were prepared as previously described (18).
Macrophages were depleted by the treatment of mice with liposomes containing dichloromethylene bisphosphonate (clodronate-liposomes), prepared as previously described (19). Clodronate was a gift from Roche (Mannheim, Germany). Macrophages were depleted in Pen V-sensitized mice by two injections of clodronate-or control PBS-liposomes; 24 h before challenge, each mouse received 0.2 ml i.p., and an additional 0.1 ml was injected i.v. 4 -6 h before elicitation of anaphylaxis. The efficacy of macrophage depletion was confirmed by flow cytometric analysis of F4/ 80 ϩ Mac-1 ϩ spleen populations.

Induction of active systemic anaphylaxis
For the induction of Pen V-induced active systemic anaphylaxis, Pen V-OVA and Pen V-BSA conjugates were prepared as previously described (20). Mice were first primed by i.p. injection of 500 g of Pen V-OVA plus 2 ϫ 10 9 Bordetella pertussis (Evans, Liverpool, U.K.; Wako Pure Chemical, Osaka, Japan) and 1.0 mg of Imject alum (Pierce, Rockford, IL). Anaphylaxis was elicited 14 days later by i.v. challenge with 100 g of Pen V-BSA (21). Rectal temperatures were measured with a digital Thermocouple thermometer linked to a murine rectal probe (Physitemp Instruments, Clifton, NJ) immediately before challenge, then every 10 min for the next 60 min. Activity levels were assessed at the same time the rectal temperatures were obtained. Clinical signs of anaphylaxis were assessed with a 0 -4 scoring system (0, normal activity; 1, mild shock including slow movement, dyspnea, ruffling of fur; 2, moderate shock including prostration sluggish gait and slight activity after prodding; 3, severe shock including complete paresis, no activity after prodding; 4, death). Commercial PAF (Calbiochem, La Jolla, CA) was used to induce systemic fatal anaphylaxis (22).

Induction of passive systemic anaphylaxis
Passive systemic anaphylaxis with immune mouse sera was induced as described previously (20). Uninfected or worm-infected BALB/c mice were immunized with 500 g of Pen V-OVA plus 2 ϫ 10 9 B. pertussis and 1.0 mg of alum. Fourteen days later, blood was recovered, and serum was extracted. To induce passive systemic anaphylaxis, naive mice were injected i.p. with 0.4 ml of anti-Pen V serum from uninfected or worminfected mice, three times at 10-min intervals. After 24 h, anaphylaxis was induced by i.v. challenge with 100 g of Pen V-BSA.
Passive IgE or IgG1 mAb-mediated systemic anaphylaxis was induced as described previously (23). Mouse anti-TNP IgG1 (200 g) or anti-TNP IgE (20 g) mAbs was administered i.v. in volumes of 200 l/mouse. Thirty minutes after injection of anti-TNP IgG1 or 24 h after injection of anti-TNP-IgE, mice were i.v. challenged with 1.0 mg of TNP 4 -OVA (Biosearch Technologies, Novato, CA) in Dulbecco's PBS. Control mice received OVA in PBS. Rectal temperatures were recorded after challenge, as described above.

In vitro worm culture
Adult worms were cultured as previously described (18). In brief, adult schistosome worms were obtained 8 wk postinfection by portal perfusion as previously described (24). The male and female worms were separated under a microscope, and male worms were washed in fresh medium. For Transwell experiments, spleen cells were plated at a density of 1 ϫ 10 7 / well in a 12-well plate (Greiner Bioscience, Frickenhausen, Germany) in 2 ml of RPMI 1640 with HEPES (25 mM), with 15 male worms placed in each Transwell. Plates were incubated for 48 h at 37°C, after which time the worms were removed, and cells were washed several times. The viability of cells was checked. Worm-modulated spleen cells were used for isolation of B cells or immunofluorescence staining.

Magnetic sorting of cells
B cells were purified by negative depletion using CD43 ϩ magnetic microbeads beads (Miltenyi Biotec, Auburn, CA). Splenic populations were incubated with anti-CD43 ϩ magnetic microbeads, followed by capture with autoMACS separation columns. The resulting B cell populations were routinely Ͼ95% pure, as assessed by FACS for CD19 expression. For transfer of purified cells, 0.5-1.0 ϫ 10 7 purified B cells were injected i.p. into individual mice 24 h before elicitation of anaphylaxis.
Sandwich ELISAs were performed to quantify levels of specific cytokines in the supernatants from cell cultures. A commercial kit (Promega, Southampton, U.K.) was used to measure total TGF-␤ (acidified samples) according to the manufacturer's instructions. Reagents for detection of IL-4 and IFN-␥ were obtained from BD Pharmingen (San Diego, CA), and IL-10 reagents were purchased as a DuoSet ELISA development system from R&D Systems (Abingdon, U.K.). Mouse mast cell protease 1 (mMCP-1) was measured in sera using a commercial kit according to the manufacturer's instructions (Moredun Scientific, Midlothian, Scotland). Histamine was measured in plasma samples using a commercial kit (IBL, Hamburg, Germany). Pen V-specific serum IgG1 and IgE levels were detected by direct ELISA (20). In brief, 96-well microtiter plates were coated with 100 l of Pen V-BSA (10 g/ml in carbonate buffer, pH 9.6) overnight at 4°C. A serial dilution of serum was added to plates that were then incubated at 4°C overnight. Detection was performed with peroxidase-conjugated IgG1 (1/1000, 2 h; Zymed Laboratories, San Francisco, CA) or biotin-labeled anti-IgE (2.5 g/ml, 2 h; BD Pharmingen), followed by peroxidase-conjugated streptavidin (1/1000, 30 min; BD Pharmingen) before substrate development using O-phenylenediamine.

Flow cytometry
Surface marker expression and intracellular cytokine phenotyping were assessed by immunofluorescent staining of B cell populations as previously described (25,26). Data were collected on a FACScan flow cytometer (BD Biosciences, San Jose, CA) and were analyzed using CellQuest software (BD Biosciences). Quadrants were drawn based on isotype-control Ig staining and were plotted on logarithmic scales.

S. mansoni infection prevents anaphylaxis
To dissect the underlying mechanism of antiallergic protective responses induced by schistosome infection, an anaphylaxis model using Pen V as the allergen was used. To investigate the relative roles of schistosome worms vs eggs in anaphylaxis, mice were infected with an S. mansoni conventional mixed male and female infection where eggs are laid (wormϩegg infection) or with a male-only infection where there are no eggs (worm-only infection). After Pen V challenge of presensitized uninfected BALB/c mice, ϳ10 -20% of animals die, with all mice having a marked drop in core body temperature within 10 min of challenge, with associated clinical signs of anaphylactic shock (Fig. 1A). Uninfected mice undergoing Pen V-induced anaphylaxis have an attendant mast cell degranulation, as determined by detection of substantial elevations in circulating levels of mMCP-1 and histamine (Fig. 1B). In contrast, BALB/c mice that had been infected with S. mansoni wormϩegg or worm-only infections were both protected from anaphylaxis, with no mortalities, reduced severity of symptoms, and negligible release of mast cell by-products (Fig. 1). Significantly, mice with worm-only infections were completely resistant to anaphylaxis, whereas wormϩegg-infected mice showed mild symptoms of anaphylaxis and had detectable, but limited (45 Ϯ 20 pg/ml), release of histamine. These results indicate that the worm stage of S. mansoni infection may elicit protection from Pen V-induced anaphylaxis.

Schistosome worm-infected mice are resistant to passive Ab-mediated anaphylaxis, but are susceptible to PAF
The reduced susceptibility of worm-infected mice to Pen V-induced anaphylaxis could be due to the infection modifying the production of anti-Pen V Ab or to the known role for helminthinduced, nonspecific, polyclonal IgE-saturating mast cells (27). However, ELISA for Pen V-specific Ab responses of sera from Pen V-OVA-immunized uninfected and worm-infected mice demonstrated that both groups had comparable titers of allergen-specific IgE and IgG1 (data not shown). Similarly, when sera from Pen V-OVA-immunized uninfected and worm-infected mice were transferred to naive animals, there were comparable levels of anaphylaxis in both groups after subsequent Pen V-BSA challenge ( Fig. 2A).
To further confirm that resistance of worm-infected mice to anaphylaxis was not due to alterations in the functionality of Ab, a FIGURE 1. S. mansoni infection protects mice from systemic anaphylaxis. Uninfected and schistosome worm-or wormϩegg-infected BALB/c mice were sensitized with Pen V-OVA on day 1. Infected mice were sensitized 7 wk after infection. Pen V-BSA was injected i.v. 14 days later to elicit anaphylaxis. Rectal body temperature and severity scores (A) were recorded before Pen V-BSA challenge (time zero) and every 10 min for 1 h. The deaths of mice are represented in brackets. Serum mMCP-1 and plasma histamine (B) levels in uninfected, wormϩegg-infected, and worm-infected (Ⅺ) or Pen V-challenged (f) mice were measured by ELISA (ND, not detected). Data are presented as the group mean Ϯ SEM (n ϭ 5-7) temperature change or severity score and are representative of three separate experiments. model of passive systemic anaphylaxis was used, in which uninfected and worm-infected mice were passively transferred IgG1 or IgE, specific for TNP, before eliciting anaphylaxis by challenge with TNP-OVA. Uninfected mice treated with TNP-IgG1 were susceptible to TNP-induced anaphylaxis, with a drop in body temperature after challenge (Fig. 2B). In contrast, the temperature of worm-infected mice was not altered by TNP-IgG1 sensitization. Similarly, worm-infected mice were not affected by anaphylaxis induced by TNP-IgE Ab treatment, which evoked a rapid and se-vere drop in temperature in uninfected mice within 5 min of challenge (Fig. 2C). Therefore, a schistosome worm infection completely protects mice from both IgG1-and IgE-mediated passive systemic anaphylaxis.
Injection of PAF induces rapid fatal anaphylaxis in mice without the need for mast cell degranulation (28). Uninfected and worminfected mice were treated with PAF to determine whether schistosome worm-infected mice were susceptible to direct PAF-induced anaphylaxis. Both uninfected and worm-infected BALB/c mice were fully susceptible to PAF, with all animals dead within 10 min of treatment (Fig. 2D). Worm-infected mice are therefore fully susceptible to direct anaphylaxis induced by PAF.
Protection from anaphylaxis is independent of CD25 ϩ T cells or F4/80 ϩ Mac-1 ϩ macrophages Previously it has been shown that regulatory cells including CD4 ϩ CD25 ϩ T cells have a suppressive role in diseases (29,30), including allergic diseases such as asthma (31). Because the levels of CD4 ϩ CD25 ϩ T cells in worm-infected mice (14 -16%) are elevated compared with those in naive uninfected (8 -11%; Fig. 3A),  we depleted CD25 ϩ cells to address the role of these regulatory T cells in worm infection and anaphylaxis (Fig. 3B). Depletion of CD25 ϩ cells had no effect on the protection from anaphylaxis afforded by worm infection (Fig. 3C). This result indicates that CD25 ϩ cells do not mediate the protection against anaphylaxis that is observed in worm-infected mice.
Because macrophages have been implicated in anaphylaxis (11,32,33) and are modulated by schistosome worm infection (18), we depleted macrophages in vivo by administration of clodronate liposomes (Fig. 4A). The depletion of macrophages had no effect on worm-mediated protection against anaphylaxis, indicating that macrophages are not an effector cell in protection from Pen Vinduced anaphylaxis (Fig. 4B).

Role for IL-10 in schistosome protection from anaphylaxis
In addition to the known increase in IL-4 production by cells from worm-infected mice (34), the levels of the regulatory cytokines IL-10 and TGF-␤ were also both elevated in worm-infected mice (Table I). Because both cytokines are implicated in modulation of the immune system by helminth parasites (8), we investigated the roles of these cytokines in the resistance of worm-infected mice to anaphylaxis. Mice received an i.p. injection of 0.5 g/mouse of anti-IL-10R or anti-TGF-␤ mAb on 2 consecutive days before anaphylaxis challenge. Sensitized uninfected BALB/c mice treated with a control mAb developed anaphylaxis, with a drop in core body temperature and mortalities (two of seven mice died) after 10 min (Fig. 5A). As shown above, worm-infected mice were protected from anaphylaxis, with no drop in core body temperature over time and no mortalities (Fig. 5A). Worm-infected mice that had been treated with a depleting TGF-␤ mAb were also completely protected from anaphylaxis (Fig. 5A), indicating that this cytokine had no role in mediating protection. In contrast, worminfected mice treated with anti-IL-10R mAb were highly susceptible to anaphylaxis, with a significant drop in core body temperature (Fig. 5A) and greater mortalities of mice (four of eight mice died) than in uninfected animals (two of seven mice died). These mAb neutralization studies suggested that IL-10, but not TGF-␤, has a role in mediating the resistance of S. mansoni worm-infected mice to anaphylaxis.
To formally validate that worm resistance to anaphylaxis was mediated by IL-10, we infected IL-10-deficient mice and their C57BL/6 wild-type controls with schistosome worms and tested their susceptibility to Pen V-induced anaphylaxis. Uninfected C57BL/6 were susceptible to anaphylaxis as determined by a drop of 3°C in body temperature (Fig. 5B). Worm-infected C57BL/6 mice were resistant to anaphylaxis, but did have a drop in temperature (Fig. 5B). As all data shown previously (Figs. 1-4) were for BALB/c strain mice, it is significant that worm infection of both BALB/c (considered Th2-biased) and C57BL/6 (considered Th1biased) strains renders mice less susceptible to anaphylaxis. Worm-infected IL-10-deficient mice developed anaphylaxis, with a marked drop in core body temperature of Ͼ5°C and high mortalities (Fig. 5B). These data support a central role for IL-10 in mediating worm-induced protection against anaphylaxis.

Role of IL-10-producing cells in worm resistance to anaphylaxis
Worm-infected mice have elevated levels of IL-10 production after spleen T cell (anti-CD3) or B cell (LPS) activation compared with uninfected mice (Table I). To empirically determine the relative roles of these defined IL-10-producing cell populations in worm resistance to anaphylaxis, we used Abs for selective depletion of each cell type in vivo (Fig. 6, A and B). Despite worm-infected mice having 2-fold more IL-10-producing CD4 ϩ T cells than uninfected mice (detected by intracellular cytokine staining; data not shown), CD4 depletion had no effect on worm-mediated protection against anaphylaxis (Fig. 6B). To assess the role of B cells in anaphylaxis, an anti-IgM Ab was administered to deplete B cells. Although Ab treatment only caused partial depletion of IgM ϩ B cells (Fig. 6A), this reduction in B cells was sufficient to render S. mansoni worm-infected mice highly susceptible to anaphylaxis, with the death of all mice after 10 min (Fig. 6C). Uninfected and a ELISA measurements of cytokine levels in the supernatants from spleen cell cultures of uninfected and worm-infected mice were performed as described in Materials and Methods. Results are presented as the mean Ϯ SD from triplicate wells. ND, not detected.
b Spleen cells from two to three uninfected and worm-infected mice were pooled. Cells were unstimulated (medium) or stimulated with plate-bound anti-CD3 (10 g/ml) or LPS (1 g/ml) for 72 h. untreated worm-infected mice showed similar trends of susceptibility and protection, respectively, to anaphylaxis as previously noted.
To specifically investigate the relative roles of B-1 vs B-2 cell subpopulations in anaphylaxis, we used mice expressing the Xid (xid) mutation that are deficient in peritoneal B-1 cells. Worminfected xid and wild-type CBA/Ca mice were both equally protected from anaphylaxis compared with the anaphylaxis observed in uninfected xid and wild-type mice (Fig. 6D). These data indicate that B-1 cells have no role in the B cell-dependent protection from anaphylaxis in worm-infected mice.

Passive transfer of in vivo or in vitro modulated B cells
The fact that IgM depletion renders worm-infected mice susceptible to anaphylaxis, and B cells from these mice produce IL-10 (Table II) suggested that B cells mediate protection. To investigate whether worm-mediated protection against anaphylaxis could be transferred by B cells, B cells were purified from the spleens of worm-infected or uninfected mice and transferred to uninfected mice. B cells from uninfected mice did not alter susceptibility to anaphylaxis (Fig. 7A). Unexpectedly, transfer of B cells from worm-infected mice not only did not protect mice from anaphylaxis, but these animals had a greater drop in body temperature,  increased mortality, and prolonged symptoms of anaphylaxis compared with mice receiving B cells from uninfected mice (Fig. 7A).

The finding that transfer of B cells from worm-infected mice exacerbates anaphylaxis does not support our initial observation that worm infection induces B cells that mediate protection.
To further determine whether B cells are involved in protection from anaphylaxis, rather than allowing 8 wk of in vivo infection to modulate the cells, an in vitro modulation system was used. We have developed a model of short term in vitro culture of distinct cell populations with live worms (18). Spleen cells from a naive mouse were cultured with 15 live worms in Transwells for 48 h at 37°C. In vitro exposure of spleen cells to worms significantly ( p Ͻ 0.05) elevated the frequency of B cells expressing IL-10 (Table II). Transfer of in vitro worm-modulated B cells made recipient mice more susceptible to anaphylaxis (Fig. 7B), data similar to what occurred with transfer of B cells modulated in vivo by worm infection (Fig. 7A). It is noteworthy that we have consistently observed that mice receiving in vivo modulated B cells (Fig. 7A) had sustained anaphylaxis, with reduced temperature until 50 min after challenge, but animals that were given in vitro worm-modulated cells rapidly recovered from anaphylaxis within 30 min (Fig. 7B).
Despite the unexpected observations that in vivo or in vitro worm-modulated B cells exacerbated anaphylaxis, both B cell populations have ϳ2-fold greater basal levels of IL-10 than uninfected or unmodulated B cells (Table II). Therefore, to confirm that IL-10-producing B cells have a role in worm-mediated protection against anaphylaxis, B cells from IL-10 Ϫ/Ϫ and naive C57BL/6 mice were modulated in vitro and transferred to Pen V-OVA immunized mice before eliciting anaphylaxis. All mice were susceptible to anaphylaxis (Fig. 7C). However, mice that received IL-10 Ϫ/Ϫ worm-modulated B cells had a markedly more severe temperature drop and a greater number of mortalities than mice receiving B cells from wild-type animals. These data support a protective role for B cell-produced IL-10 in anaphylaxis.

Exogenous IL-10 or IL-4 alters susceptibility of mice to anaphylaxis
Our initial data from depletion studies suggested that schistosome worm infection induces B cells producing elevated levels of IL-10 that render the mice refractory to experimental anaphylaxis. Unexpectedly, transfer of B cells from worm-infected mice or in vitro worm-modulated cells to naive mice not only did not mediate protection from anaphylaxis, but significantly exacerbated disease (Fig. 7). However, worm-infected mice not only have elevated numbers of IL-10-producing B cells, but by intracellular cytokine staining these mice also have 4-to 5-fold elevations in the frequency of IL-4-producing B cells compared with uninfected mice (Table II and Fig. 8A). The elevation in both IL-10 and IL-4 in B cells was confirmed by RT-PCR of purified B cells isolated from both wormϩegg-infected and worm-infected mice (Fig. 8, B and C). Therefore, in the experiments described in Fig. 7, we were not only injecting mice with IL-10-producing B cells, but we were also transferring IL-4-producing B cells. Indeed, the greater ratio of IL-4-producing:IL-10-producing B cells in worm-infected mice (ϳ8:1) compared with that in in vitro modulated B cells (ϳ3:1) could explain the more severe anaphylaxis when B cells from infected mice were transferred into naive animals (Fig. 7A) compared with that after transfer of in vitro modulated B cells (Fig.  7B). Because IL-4 is known to exacerbate experimental anaphylaxis (35), the increased anaphylaxis in mice receiving B cells from worm-infected mice or in vitro worm-modulated B cells could be due to cotransfer of these IL-4-producing cells. To address this scenario, we first formally examined the relative roles of IL-4 and IL-10 in the Pen V-mediated anaphylaxis model. Sensitized mice were administered rIL-4 with or without rIL-10 before Pen V-BSA challenge. Analysis of the serum levels of IL-4 and IL-10 in these mice (Fig. 9A) demonstrated that there were artificially elevated levels of each cytokine in mice during challenge compared with control mice. Administration of rIL-4 exacerbated anaphylaxis in recipient mice, with these mice showing a 2°C greater drop in body temperature compared with control mice and a prolonged lapse in recovery (Fig. 9B). In contrast, mice that received rIL-10 were protected from anaphylaxis, with only a slight drop (ϳ1°C) in body temperature at 10 min, followed by a rapid recovery (Fig.  9C). Interestingly, mice that were treated with both rIL-4 and rIL-10 showed a similar trend as mice that received rIL-10. These in vivo data demonstrate that IL-10 can reduce the deleterious effects of IL-4 in anaphylaxis.

Transfer of in vitro worm-modulated, Th2 cytokine-deficient B cells
To exclude a role for IL-4 and other Th2 cytokines (IL-5, IL-9, and IL-13) known to induce allergic responses, a B cell transfer experiment was performed with cells from multiknockout (mKO) mice that have a combined deficiency in IL-4, IL-5, IL-9, and IL-13 (15). B cells from mKO and wild-type mice were modulated with worms in vitro and transferred to Pen V-OVA-immunized mice before Pen V-BSA challenge and induction of anaphylaxis. In vitro modulation of mKO splenocytes with worms increased the frequency of IL-10-producing B cells by 32% (data not shown). 14 c a Cells were gated on CD19 expression and quadrants for IL-4, IFN-␥, or IL-10 set with isotype control mAbs. Data presented are the mean percentage Ϯ SE of cytokine-producing cells from four to six separate experiments.
b Spleen cells were isolated from two to three uninfected and S. mansoni worm-or wormϩegg-infected mice and pooled. Cells were cultured in vitro with brefeldin A (10 g/ml) for 3 h. c p Ͻ 0.05 or d p Ͻ 0.01 significant elevation in cytokine-producing cells in schistsome-modulated B cells relative to unmodulated cells. Statistical analysis was performed using the unpaired t test with Welch correction. e Spleen cells from naive mice were unmodulated or modulated with 15 male worms for 48 h at 37°C. Cells were cultured in vitro with brefeldin A (10 g/ml) for 3 h.
Mice that received unmodulated B cells from mKO or cells from wild-type animals developed anaphylaxis (Fig. 10A). Similar to earlier data (Fig. 7B), mice that received worm-modulated B cells from wild-type mice had a more severe initial (first 10 min) drop in temperature, after which they recovered rapidly (Fig. 10A). However, passive transfer of worm-modulated B cells from mKO mice completely protected animals from anaphylaxis (Fig. 10).
To formally determine the roles of the individual Th2 cytokines, spleen cells from IL-4-, IL-5-, IL-9-, or IL-13-deficient mice were modulated by worms in vitro, and isolated B cells were transferred to Pen V-sensitized mice. Worm-modulated B cells from mice deficient in IL-4 completely protected recipients from anaphylaxis (Fig. 10B). In contrast, animals receiving modulated B cells from mice deficient in IL-5, IL-9, or IL-13 developed a comparable drop in temperature, as seen in animals administered wild-type cells. These results further corroborate a protective role for schistosome worm-modulated B cells in preventing anaphylaxis in the absence of IL-4.

Discussion
In this study we demonstrate that mice infected with S. mansoni are refractory to experimental systemic anaphylaxis. It was established that it is the worm stage of infection that elicits the protective phenotype in this model, with worm-infected mice completely protected against anaphylaxis, whereas wormϩegg-infected mice were only partially protected. We have identified that schistosome worm infection of mice prevents anaphylaxis via a B cell-and IL-10-dependent mechanism. FIGURE 7. Transfer of B cells from worm-infected mice or B cells modulated in vitro with worms does not prevent anaphylaxis in recipient mice, and this effect is exacerbated in the absence of IL-10. A, Transfer of B cells from worm-infected mice to Pen V-sensitized mice induces pronounced anaphylaxis. CD19 ϩ B cells were isolated from spleen cells from uninfected and worm-infected mice and injected (total of 1 ϫ 10 7 cells) into Pen V-sensitized BALB/c mice 2 days and 1 day before inducing anaphylaxis. Data are presented as the group mean Ϯ SEM (n ϭ 5-7) temperature change and are representative of three separate experiments. Deaths of mice are represented in brackets. B, Transfer of B cells exposed to worms in vitro to Pen V-sensitized mice transiently enhances anaphylaxis. B cells were isolated from spleen cells from BALB/c mice that had been modulated by worms in vitro (mod) or unmodulated (unmod). C, Worm-modulated B cells deficient in IL-10 exacerbate anaphylaxis when transferred to Pen V-sensitized mice. Spleen cells from an IL-10-deficient mouse (IL-10 KO) and a wild-type C57BL/6 mouse were worm-modulated (mod) or umodulated (unmod) in vitro. B cells were isolated and transferred to Pen V-sensitized C57BL/6 mice as described above. Deaths of mice are represented in brackets. Data are presented as the group mean Ϯ SEM (n ϭ 5-8) temperature change and are representative of three separate experiments. The central role of B cells in protecting worm-infected mice from anaphylaxis was illustrated by the finding that even partial depletion of B cells caused these animals to be fully susceptible to anaphylaxis. B cells possess a variety of immune functions, including production of Igs, presentation of Ags, and production of distinct cytokines (36). B cells have also been shown to participate in the induction of immune tolerance and suppression of disease, giving rise to the suggestion of IL-10-producing regulatory B cells (37)(38)(39). Previously, it has indeed been shown that B cells play an important regulatory role during conventional S. mansoni wormϩegg infection, with high mortalities of infected MT mice (40). We show in this study that B cells from schistosome worminfected mice have 2-fold elevated levels of IL-10-producing B cells, which is consistent with worm-mediated protection from anaphylaxis being dependent on IL-10 and B cells. Interestingly, wormϩegg-infected mice also had significantly elevated IL-10-producing B cells. It is well documented that schistosome wormϩegg infections induce elevated IL-10 production (41-43). IL-10 production during such infections has an essential function in suppressing egg-induced immunopathology (44 -47). Previously, peritoneal B-1 cells from wormϩegg-infected mice have been shown to produce IL-10 (48). Therefore, to determine whether B-1 cells were involved in worm infection-induced resistance to anaphylaxis, we infected B-1 cell-deficient xid mice. Worm-infected, B-1 cell-deficient mice were resistant to Pen V-induced anaphylaxis, indicating that a B-2 cell population mediates resistance from anaphylaxis. Our future studies will address whether a distinct B-2 cell subpopulation preferentially produces IL-10.
In this study spleen cells from schistosome worm-infected mice also have elevated IL-10 production, with resistance of these mice to anaphylaxis being IL-10 dependent. IL-10 is one of the main cytokines that is implicated in certain regulatory T cell types (49) and, with respect to asthma, has also been shown to suppress immune and inflammatory responses, including allergic airway inflammation (50). In murine schistosome wormϩegg infections, a recent study has addressed for the first time the role of IL-10producing CD4 ϩ CD25 ϩ T cell in ameliorating morbidity (47). However, in this study, although natural CD4 ϩ CD25 ϩ T cells and IL-10-producing T cells are both elevated in worm-infected mice, depletion studies demonstrated that CD4 ϩ or CD25 ϩ cells had no role in worm-mediated protection against anaphylaxis. The data presented here indicate that T cells are not the only source of IL-10 during a schistosome worm infection, an observation also demonstrated during S. mansoni wormϩegg infections by Hesse et al. (47). Because schistosome infection of mice stimulates a number  of cells to produce IL-10, it highlights the known significant role for this cytokine in immunomodulation of the host to suppress parasite-induced pathology.
Previously, a protective role for IL-10-producing B cells has been demonstrated in experimental encephalomyelitis and collagen-induced arthritis (26,37). In experimental encephalomyelitis, IL-10-producing B cells change the recovery of mice from disease through suppression of type 1 autoantigen responses (37). Similarly, transfer of in vitro anti-CD40-activated B cells from spleens of arthritic DBA/1 mice prevented collagen-induced arthritis in recipients, with B cells from IL-10 Ϫ/Ϫ mice failing to mediate this protective function (26). In this study transfer of B cells from schistosome worm-infected mice (in vivo modulated) or in vitro worm-modulated B cells not only did not prevent anaphylaxis, as we expected, but induced more severe anaphylaxis in recipients. Because we attributed this failure to confer protection to cotransfer of IL-4-producing B cells (Fig. 8), it would explain the discrepancy with B cell transfer protecting in the collagen-induced arthritis (26), because in this model IL-4 is protective; treatment of mice with engineered IL-4-producing B cells (51) or IL-4-expressing dendritic cells (52) was shown to reduce or suppress arthritis.
During cytokine phenotyping of B cells in infected mice we had observed that splenocytes from schistosome worm-infected mice had 4-to 5-fold more IL-4-producing B cells than those from uninfected mice. Because schistosome worm infections are type 2 cytokine-biased (34), it is perhaps not unexpected that B cells from worm-infected mice have a type 2, or B effector 2 (Be2), profile. In the more Th2-biased wormϩegg-infected mice, there was an even more marked increase in levels of IL-4-producing cells (Table II and Fig. 8). It has previously been shown that mesenteric lymph node cells from mice infected with the Th2-inducing nematode worm Heligmosomoides polygyrus have elevated Be2 cells (36). Thus, upon transferring worm-modulated B cells that had elevated IL-10 to recipient mice, we were also transferring IL-4producing Be2 cells. This contamination with Be2 cells became relevant after Finkelman and colleagues (35) demonstrated that injection of mice with only 65 ng of IL-4 significantly increased the severity of anaphylaxis. Indeed, earlier findings by Park et al. (20) also demonstrated that blockade of IL-4 by anti-IL-4 mAb treatment prevented fatal anaphylaxis in the same antibiotic model as that used in this study. We therefore treated mice with exogenous IL-4 and/or IL-10 and not only illustrated the respective deleterious or protective effects of these cytokines in Pen V-induced anaphylaxis, but also showed that IL-10 can suppress the pathogenic effects of IL-4. Previously, it was reported that treatment of mice with 50 g of IL-10 had no effect on their susceptibility to anaphylaxis (35), whereas we observed that administration of 1 g of IL-10 protected mice. Because this discrepancy could be due to the difference in anaphylaxis models used, it would be of interest to evaluate the effects of IL-10 in different models of anaphylaxis. The use of mice deficient in Th2 cytokines permitted confirmation that it was up-regulation of IL-4 in schistosome worm-modulated B cells that was responsible for the exacerbated anaphylaxis in recipient mice given worm-modulated B cells. Therefore, in the experiments described in Fig. 7, in which transferred B cells exacerbated anaphylaxis, we were not only injecting mice with IL-10-producing B cells, but we were also transferring IL-4-producing B cells. In the B cell transfer studies, mice injected with in vivo modulated cells had more severe anaphylaxis than mice receiving in vitro modulated cells, which is consistent with B cells from infected mice having greater relative frequencies of IL-4-producing:IL-10-producing B cells.
Bashir et al. (53) have demonstrated a role for experimental helminth infection in protection against the development of allergy in mice. They assessed the effect of Th2 responses induced by the intestinal helminth infection (H. polygyrus) on the development of an allergic response to the food allergen peanut Ag, with protection being mediated at least in part by the production of IL-10. Although we have demonstrated a novel mechanism by which a helminth parasite can prevent anaphylaxis in mice, it is possibly too simplistic to envisage a common mechanism by which different helminth parasites may prevent allergic responses. This can be illustrated by infection of mice with a gastrointestinal nematode (H. polygyrus) that has been shown to reduce allergic responses (53) in one study, whereas in a separate study mice infected with another gastrointestinal nematode (Trichinella spiralis) had exacerbated anaphylaxis (35). There are defined differences in the infectivity and immunity of these two parasites (54) as well as in the allergy models used that could explain the differences between the studies.
In the context of the hygiene hypothesis, it has previously been proposed that schistosome infection can ameliorate allergic disorders in humans (8). In support of this argument, schistosome-infected Gabonese schoolchildren have fewer allergic responses and elevated IL-10 levels (55), with an increase in atopic responses in the schoolchildren after chemotherapy to eliminate helminth infections (56). Using experimental animal studies, we now formally demonstrate that S. mansoni worm infection of mice can protect mice from anaphylaxis via an immunoregulatory response. Using different anaphylaxis models, it is evident that worm infection does not alter Ag sensitization, but protects the mice from anaphylaxis when they are rechallenged with the allergen. Therefore, resistance of schistosome-infected mice to anaphylaxis was not due to the known propensity of helminth infections to alter responses to vaccination (57) or the stimulation of polyclonal IgE, which can saturate Fc⑀Rs and thereby prevent mast cell degranulation (27). Protection from anaphylaxis is thus primarily mediated when the sensitized mice are rechallenged with the appropriate allergen.
It is noteworthy that when a protective factor, i.e., B cells or IL-10, was inhibited, the worm-infected mice had a heightened propensity to develop fatal anaphylactic reactions upon Ag challenge. This highlights that in the absence of regulation by IL-10producing B cells, schistosome-infected mice are, in fact, highly predisposed to type 2 allergic responses, thereby illustrating the sensitive balance of S. mansoni modulation of host immunity to achieve host protective vs detrimental immune responses. Hence, IL-10-producing B cells are necessary to maintain a fine balance of regulation in the allergic-predisposed environment of schistosome infection.
Collectively, these experimental data demonstrate that schistosome worms evoke IL-10-producing B cells during infection, with these cells protecting against anaphylaxis. We show the IL-10producing B cells are induced during schistosome infection as part of the parasite's regulation of host immunity, with this modulation also influencing unrelated allergic (anaphylaxis) responses. These data offer for the first time conclusive experimental evidence that S. mansoni can suppress allergic responses via a novel regulatory cell mechanism.